ClinCapture

Here is a fact about clinical trials that is either obvious or slightly deranged, depending on how close you sit to one: the protocol - the document that says what a trial will do, to whom, when, and how the data gets recorded - starts life as prose. Someone writes it in a word processor. Then someone else reads that prose and, by hand, configures a database to match it. Then a third person checks whether the database actually matches the prose. Errors, being errors, arrive somewhere in the handoff. This is the machinery that decides whether a drug works, and a meaningful chunk of it is a person retyping a Word document into a form builder.

ClinCapture sells the machinery. It is an eClinical software company - a phrase that means "the software clinical trials run on" - and it has been quietly doing so since 2010, when Ale Gicqueau spun it out of a Silicon Valley clinical-data consultancy called Clinovo. The company is headquartered in Las Vegas, employs somewhere around two dozen people, and counts among its customers a client list that would look padded if it weren't apparently real: Alcon, Abbott, Teleflex, LabCorp, Hologic, AstraZeneca, W.L. Gore, Edwards Lifesciences, Biogen. These are companies that run trials for a living, and they are notoriously hard to sell to, because the cost of a software vendor being wrong is not a refund - it's a regulatory finding.

The thing that makes ClinCapture interesting is not that it built another electronic data capture system. Plenty of people have. It's how it got in the door. For years the company's headline pitch was that ClinCapture was the only free, fully validated EDC in the world - open-source software, given away, in the single most compliance-allergic industry on the planet. That is the sort of sentence that makes a chief financial officer's eye twitch. Free software in drug development sounds like a category error. It worked anyway, and understanding why is most of the story.

What it actually does

Strip away the acronyms and a clinical trial is a data-collection problem wearing a lab coat. You need to enroll patients (and prove they consented), capture what happens to them (from clinics and increasingly from their own phones), randomize who gets what, keep the drug supply straight, translate every reported symptom into a standardized medical vocabulary, and file every document so that when a regulator asks "show me," you can. Historically each of those jobs came from a different vendor, in a different login, speaking a different data format. The polite industry term for the result is "point solutions." The impolite term is "a mess."

ClinCapture's answer is a single platform - now branded Captivate - that tries to do all of it in one place. Electronic data capture with no-code study design. Virtual Data Capture for patient-reported outcomes and electronic consent, aimed at decentralized and hybrid trials where the patient never has to drive to a site. Randomization and trial supply management. Medical coding with a full audit trail. An electronic trial master file so the document pile is inspection-ready rather than inspection-adjacent. The pitch is not "our EDC is better." The pitch is "stop stitching."

The AI turn, and why the framing matters

In March 2026 the company did the thing every software company did in 2026 - it announced AI - but the framing is worth pausing on, because ClinCapture went out of its way to say it was doing the opposite of what everyone else was doing. The fashionable move is to bolt an AI agent on top of an existing product: a chatbot that reads your data, a copilot in the corner. ClinCapture's claim is that it embedded AI into the architecture of the trial itself - an engine that takes a structured protocol and automatically generates and configures large portions of the study, inside the validated environment, before a single patient is enrolled.

CEO Scott Weidley calls the goal "regulated intelligence," which is a genuinely useful phrase for anyone shipping AI into a regulated field: AI that amplifies domain expertise without compromising validation, auditability, or quality. In an industry where "the model made a mistake" can mean "the trial is compromised," the interesting engineering is not the model. It's the guardrails around it - the ISO 27001 and ISO 9001 certifications, the 21 CFR Part 11 and GDPR compliance, the audit trails that let a human stay in the loop and a regulator stay satisfied. ClinCapture's bet is that the boring parts are the product.

A 23-person company does not usually get to name-drop like this. ClinCapture's publicly cited clients skew heavily toward medical devices and pharma - the exact customers who need validation more than they need charm.